Cardiovascular diseases accounted for approximately 43 percent of the mortality in the United States of America in 1991 (923,000 persons). However, many of these deaths are not directly caused by an acute myocardial infraction (AMI). Rather, many patients suffer a general decline in their cardiac output known as heart failure. Once the overt signs of heart failure appear, half the patients die within five years. It is estimated that between two and three million Americans suffer from heart failure and an estimated 200,000 new cases appear every year. In many cases heart failure is caused by damage accumulated in the patient's heart, such as damage caused by disease, chronic and acute ischemia and especially (˜75%) as a result of hypertension.
A short discussion of the operation of a healthy heart is usefull in order to appreciate the complexity of the functioning of the heart and the multitude of pathologies which can cause heart failure. FIG. 1A is a schematic drawing of a cross-section of a healthy heart 20. In general heart 20 comprises two independent pumps. One pump comprises a right atrium 22 and a right ventricle 24 which pump venous blood from an inferior and a superior vena cava to a pair of lungs (not shown) to be oxygenated. Another pump comprises a left atrium 26 and a left ventricle 28, which pump blood from pulmonary veins (not shown) to a plurality of body systems, including heart 20 itself. The two ventricles are separated by a ventricular septum 30 and the two atria are separated by an atrial septum 32.
Heart 20 has a four phase operational cycle in which the two pumps are activated synchronously. FIG. 1B shows a first phase, called systolic. During this phase, right ventricle 24 contracts and ejects blood through a pulmonic valve 34 to the lungs. At the same time, left ventricle 28 contracts and ejects blood through an aortic valve 36 and into an aorta 38. Right atrium 22 and left atrium 26 are relaxed at this point and they begin filling with blood, however, this preliminary filling is limited by distortion of the atria which is caused by the contraction of the ventricles.
FIG. 1C shows a second phase, called rapid filling phase and indicates the start of a diastole. During this phase, right ventricle 24 relaxes and fills with blood flowing from right atrium 22 through a tricuspid valve 40, which is open during this phase. Pulmonic valve 34 is closed, so that no blood leaves right ventricle 24 during this phase. Left ventricle 28 also relaxes and is filled with blood flowing from left atrium 26 through a mitral valve 42, which is open. Aortic valve 36 is also closed to prevent blood from leaving left ventricle 26 during this phase. The filling of the two ventricles during this phase is affected by an existing venous pressure. Right atrium 22 and left atrium 26 also begin filling during this phase. However, due to relaxation of the ventricles, their pressure is lower than the pressure in the atria, so tricuspid valve 40 and mitral valve 42 stay open and blood flows from the atria into the ventricles.
FIG. 1D shows a third phase called diastatis, which indicates the middle of the diastole. During this phase, the ventricles fill very slowly. The slowdown in filling rate is due to the equalization of pressure between the venous pressure and the intra-cardiac pressure. In addition, the pressure gradient between the atria and the ventricles is also reduced.
FIG. 1E shows a fourth phase called atrial systole which indicates the end of the diastole and the start of the systole of the atria During this phase, the atria contract and inject blood into the ventricles. Although there are no valves guarding the veins entering the atria, there are some mechanisms to prevent backflow during atrial systole. In left atrium 26, sleeves of atrial muscle extend for one or two centimeters along the pulmonary veins and tend to exert a sphincter-like effect on the veins. In right atrium 22, a crescentic valve forms a rudimentary valve called the eustachian valve which covers the inferior vena cava. In addition, there may be muscular bands which surround the vena cava veins at their entrance to right atria 22.
FIG. 1F is graph showing the volume of left ventricle 24 as a function of the cardiac cycle. FIG. 1F clearly shows the additional volume of blood injected into the ventricles by the atria during atrial systole as well as the variance of the heart volume during a normal cardiac cycle. FIG. 1G is a graph which shows the time derivative of FIG. 1F, i.e., the left ventricle fill rate as a function of cardiac cycle. In FIG. 1G two peak fill rates are shown, one in the beginning of diastole and the other during atrial systole.
An important timing consideration in the cardiac cycle is that the atrial systole must complete before the ventricular systole begins. If there is an overlap between the atrial and ventricular systoles, the atria will have to force blood into the ventricle against a raising pressure, which reduces the volume of injected blood. In some pathological and induced cases, described below, the atrial systole is not synchronized to the ventricular systole, with the effect of a lower than optimal cardiac output.
It should be noted that even though the left and the right sides of heart 20 operate in synchronization with each other, their phases do not exactly overlap. In general, right atrial systole starts slightly before left atrial systole and left ventricular systole starts slightly before right ventricular systole. Moreover, the injection of blood from left ventricle 26 into aorta 38 usually begins slightly after the start of injection of blood from right ventricle 24 towards the lungs and ends slightly before end of injection of blood from right ventricle 24. This is caused by pressures differences between the pulmonary and body circulatory systems.
When heart 20 contracts (during systole), the ventricle does not contract in a linear fashion, such as shortening of one dimension or in a radial fashion. Rather, the change in the shape of the ventricle is progressive along its length and involves a twisting effect which tends to squeeze out more blood. FIG. 2 shows an arrangement of a plurality of muscle fibers 44 around left ventricle 28 which enables this type of contraction. When muscle fibers 44 are arranged in a spiral manner as shown in FIG. 2 and the activation of muscle fibers 44 is started from an apex 46 of left ventricle 28, left ventricle 28 is progressively reduced in volume from the bottom up. The spiral arrangement of muscle fibers 44 is important because muscle fibers typically contract no more than 50% in length. A spiral arrangement results in a greater change of left ventricular volume than is possible with, for example, a flat arrangement in which the fibers are arranged in bands around the heart. An additional benefit of the spiral arrangement is a leverage effect. In a flat arrangement, a contraction of 10% of a muscle fiber translates into a reduction of 10% of the ventricular radius. In a spiral arrangement with, for example, a spiral angle 48 of 45°, a 10% contraction translates into a 7.07% contraction in ventricular radius and a 7.07% reduction in ventricular length, Since the ventricular radius is typically smaller than the ventricular length, the net result is that, depending on spiral angle 48, a tradeoff is effected between a given amount of contraction and the amount of force exerted by that contraction.
Spiral angle 48 is not constant, rather, spiral angle 48 changes with the distance of a muscle fiber from the outer wall of the ventricle. The amount of force produced by a muscle fiber is a function of its contraction, thus, each layer is optimized to produce an optimal amount of force. Since the contraction of each muscle fiber is synchronous with the increase in the ventricular pressure (caused by the muscle contraction), it might be expected that the muscle fibers produce a maximum force at maximum contraction. However, physiological constraints on muscle fibers denote that maximal force is generated before maximal contraction. In addition, the force exerted by a muscle fiber begins to fall soon after maximum force is exerted. The varying spiral angle is a mechanism which makes it possible to increase the contractile force on the ventricle after maximum force is reached by a particular muscle fiber.
As described above, activation of the heart muscle is from the apex up. Thus, the muscle on the top of the ventricle could theoretically exert more force than the muscle at apex 46, which would cause a distention at apex 46. The varying spiral angle is one mechanism to avoid distention. Another mechanism is that the muscle near apex 46, which is activated first, is slightly more developed than the muscle at the top of the ventricle, which is activated last. As a result of the above described mechanisms, the force exerted by the ventricular wall is more evenly distributed over time and space. It should be appreciated that blood which remains in one place without moving, even in the heart, can clot, so it is very important to eject as much blood as possible out of the heart.
As can be appreciated, a complicated mechanism is required to synchronize the activation of muscle fibers 44 so that an efficient four phase cycle is achieved. This synchronization mechanism is provided by an electrical conduction system within the heart which conducts an electrical activation signal from a (natural) cardiac pacemaker to muscle fibers 44.
FIG. 3 shows the main conduction pathways in heart 20. An SA node 50, located in right atrium 22, generates an activation signal for initiating contraction of muscle fibers 44. The activation signal is transmitted along a conduction pathway 54 to left atria 26 where the activation signal is locally disseminated via Bachman bundles and Crista terminals. The activation signal for contracting the left and right ventricles is conducted from SA node 50 to an AV node 52, where the activation signal is delayed. The ventricles are normally electrically insulated from the atria by non-conducting fibrous tissue, so the activation signal must travel through special conduction pathways. A left ventricle activation signal travels along a left pathway 58 to activate left ventricle 28 and a right ventricle activation signal travels along a right pathway 56 to activate right ventricle 24. Generally, the conduction pathways convey the activation signal to apex 46 where they are locally disseminated via Purkinje fibers 60 and propagation over the rest of the heart is achieved by conduction in muscle fibers 44. In general, the activation of the heart is from the inner surface towards the outer surface. It should be noted that electrical conduction in muscle fibers 44 is generally faster along the direction of the muscle fibers. Thus, the conduction velocity of the activation signals in heart 20 is generally anisotropic.
As can be appreciated, the delay in AV node 52 results, in a healthy heart, in proper ventricular systolic sequencing. The temporal distribution of the activation signal in the ventricular muscle results in the activation of the ventricles from the apex up. In a healthy heart the activation signal propagates across left ventricle 28 in approximately 60 milliseconds. In an externally paced heart, where the activation signal is not conducted through Purkinje fibers 60 or in a diseased heart, the propagation time is typically longer, such as 150 milliseconds. Thus, disease and external pacing affect the activation profile of the heart.
Cardiac muscle cells usually exhibit a binary reaction to an activation signal; either the cell responds normally to the activation signal or it does not respond at all. FIG. 4 is a graph showing changes in the voltage of a single cardiac muscle cell in reaction to the activation signal. The reaction is generally divided into five stages. A rapid depolarization stage 62 occurs when the muscle cell receives an activation signal. During this stage, which lasts a few milliseconds, the potential of the cell becomes rapidly positive. After depolarization, the muscle fiber rapidly repolarizes during a rapid repolarization stage 64 until the cell voltage is approximately zero. During a slow repolarization stage 66, also known as the plateau, the muscle cell contracts. The duration of stage 66, the plateau duration, is directly related to the amount of work performed by the muscle cell. A relatively fast repolarization stage 68 follows, where the muscle cell repolarizes to its original potential. Stage 66 is also known as the refractory period, during which the cell cannot be activated by another activation signal. During stage 68, the cell is in a relative refractory period, during which the cell can be activated by an exceptionally strong activation signal. A steady state 70 follows in which the muscle cell is ready for another activation.
It should be appreciated that the contraction of cardiac muscle cells is delayed in time from their activation. In addition the duration of the contraction is generally equal to the duration of the plateau.
An important factor which may affect the length of the plateau is the existence of an ionic current resulting from the voltage potentials generated by the local depolarizations. The ionic current starts at the last activated portion of the heart and progresses back along the path of the activation. Thus, it is the later activated portions of the heart which are first affected by the ionic current. As a result, the repolarization of these cells is relatively faster than the repolarization of the first activated muscle fibers, and their contraction time is relatively shorter. As can be appreciated, in a healthy heart, where the propagation time of the activation signal is relatively short, the ionic currents are significantly smaller than in a diseased or externally paced heart.
One of the main results of the contraction of the ventricles is increased intra-ventricular pressure. In general, when the intra-cardiac pressure is higher, the outflow from the heart into the circulatory system is stronger and the efficiency of the heart is higher. A mathematical relationship termed Laplace's law can be used to model the relationship between the pressure in the ventricle and the tension in the wall of the ventricle. Laplace's law was formulated for generally spherical or cylindrical chambers with a distentible wall, however, the law can be applied to the ventricles since they are generally elongated spherical in shape. FIGS. 5A–C show three formulations for determining the tension in a portion of the ventricle wall, all of which are based of the law of Laplace. In FIG. 5A, the tension across a cross-section of the wall is shown wherein T, the tension in the wall, is equal to the product of P, the transmural pressure across the wall, r (squared), the radius of the ventricle, and π. FIGS. 5B and C show formulas for calculating the tension per unit in portions of the ventricular wall, for example in FIG. 5C, for a unit cross-sectional area of muscle in a wall of thickness δ.
As can be appreciated, if r, the radius of the ventricle, is large, a higher tension is needed to produce the same pressure change as in a ventricle with a smaller radius. This is one of the reasons that ventricular dilation usually leads to heart failure. The heart muscle is required to produce a higher tension is order to achieve the same pressure gradient. However, the heart is not capable of producing the required tension, so, the pressure gradient, and thus the cardiac efficiency, are reduced.
Unfortunately, not all people have healthy hearts and vascular systems. Some types of heart problems are caused by disease. HCM (hypertrophic cardiomyopathy or HOCM) is a disease in which the left ventricle and, in particular, the ventricular septum, hypertrophy, sometimes to an extent which blocks the aortic exit from the left ventricle. Other diseases, such as atrophy causing diseases, reduce the amount of muscle fibers in portions of the heart.
A very common cause of damage to the heart is ischemia of the heart muscle. This condition, especially when manifesting itself as an acute myocardial infraction (heart attack), can create dead zones in the heart which do not contain active muscle. An additional, and possibly more important effect, is the non-conducting nature of these dead zones which may upset the natural activation sequence of the heart. In some cases, damaged heart tissue continues to conduct the activation signal, albeit at a variable or lower velocity, which may cause arrhythmias.
A chronic ischemic condition is usually caused by blockage of the coronary arteries, usually by arteriosclerosis, which limits the amount of oxygen which can reach portions of the heart muscle. When more work (i.e., more tension) is required of the heart muscle and an increase in oxygen supply is not available, the result is acute pain, and if the supply is cut off for an extended period, death of the starved muscle will follow.
When the output of the heart is insufficient, a common result is hypertrophy of the heart, usually of the left ventricle. Hypertrophy is a compensatory mechanism of the heart for increasing the output volume. However, in a chronic condition, hypertrophy has generally negative effects. For example, arrhythmias, congestive heart failure (CHF) and permanent changes in the morphology of the heart muscle (ventricular modeling) may result from hypertrophy.
One of the most common cardiovascular diseases is hypertension. A main effect of hypertension is increased cardiac output demand, which causes hypertrophy since the blood must be pumped against a higher pressure. Furthermore, hypertension usually aggravates other existing cardiac problems.
The human heart has many compensatory and adaptive mechanisms, termed cardiac reserve, so that not all cardiac pathologies manifest as heart disease. Once the cardiac reserve is used up, the heart cannot keep up with the demand and heart failure may result. One measure of heart function and efficiency is the left ventricle ejection factor, which is the ratio between the amount of blood in the left ventricle during diastole and the amount of blood exiting during systole. It should be noted that a significant portion of the change in ventricular volume between systole and diastole is due to the thickening of activated muscle fibers. Another measure of heart function is the left ventricle stroke volume, which is the amount of blood which is ejected from the left ventricle each heart beat. It should be noted that once the cardiac reserve is used up it is difficult, if not impossible, for the heart to increase its output when needed, such as during exercise.
There are many-ways in which non-optimal timing of the activation of the heart can result in lower cardiac output. In AF (atrial fibrillation) one or both atria does not contract in correct sequence with its associated ventricle. As a first result, the atria does not inject blood into its associated ventricle during atrial systole, so the ventricle volume is not maximized before ventricular systole, and stroke volume is slightly reduced. If the right atria is fibrillating, sequencing of the AV node is non-regular, which results in the ventricles contracting at an irregular rate, and the heart output is further reduced.
In some cases of a conduction block between the SA node and the ventricles, such as caused by a damaged AV node, the contraction of the atria is not synchronized to the contraction of the ventricles, which also results in a lower heart output.
Another type of timing deficiency results when there are large dead areas in the heart muscle which do not conduct electrical signals. The activation signal must circumvent the dead areas, which results in a longer pathway (and longer delay time) for the activation signal reaching some portions of the heart. In some cases, these portions of the heart are activated long after the rest of the heart has already contracted, which results in a reduced contribution of these portions to the total cardiac output.
Heart muscle which is stressed before it is activated, heart muscle which is weakened (such as by ischemia) and portions of the heart which have turned into scar tissue, may form aneurysms. As can be appreciated from Laplace's law, portions of the ventricle wall which do not generate enough tension to offset the tension induced by the intra-cardiac pressure must increase their local radius in response to the pressure overload. The stretched wall portion thins out and may burst, resulting in the death of the patient. The apex of the left ventricle is especially susceptible to aneurysms since it may be very thin. In addition, the total pressure in the ventricle and the flow from the ventricle are reduced as the aneurysm grows, so the heart output is also reduced. Although weak muscle should be expected to hypertrophy in response to the greater need, in some cases, such as after an AMI, hypertrophy may not occur before irreversible tissue changes are caused by the stretching.
Perfusion of the heart muscle usually occurs during diastole. However, if the diastole is very long, such as when the activation signal is propagated slowly, some portions of the heart may not be oxygenated properly, resulting in functional ischemia.
As mentioned above, one of the adaptation mechanisms of the heart is hypertrophy, in which the size of the heart increases to answer increased demand. However, hypertrophy increases the danger of arrhythmias, which in some cases reduce heart output and in others, such as VF (ventricular fibrillation) are life threatening. Arrhythmias are also caused by damaged heart tissues which generate erroneous activation signals and by blocks in the conduction system of the heart.
In some cases arrhythmias of the heart are treated using medicines, in others, by implanting a pacemaker or a defibrillator. A common pacemaker implanting procedure, for example for treating the effects of AF, includes:
(a) ablating or removing the AV node; an
(b) implanting a pacing electrode in the apex of the heart. The location of the pacing electrode may be changed (during the procedure) if the heart does not beat at a desired sequence for a given output of the pacemaker.
It is also known to pace using multiple electrodes, where the activation signal is initiated from a selected one or more of the electrodes, depending on sensed electrical values, such as sequence, activation time and depolarization state. Typically, the pacing regime is adapted to a specific arrhythmia. Sometimes, logic is included in the pacemaker which enables it to identify and respond to several types of arrhythmia.
U.S. Pat. No. 5,403,356 to Hill et al. describes a method of preventing atrial arrhythmias by adapting the pacing in the right atrium in response to a sensed atrial depolarization, which may indicate an arrhythmia.
Sometimes the pacing is performed for more than one chamber. For example, in dual chamber pacing, both left and right ventricles are separately paced. There have been attempts to use dual chamber pacing to relive aortic obstruction caused by HCM. The aortic exit from the left ventricle is located between the left and right ventricle, so that when both ventricles contract simultaneously, the aorta is squeezed from all sides. In a healthy heart, the ventricular septum does not obstruct the aorta, however, in an HCM-diseased heart, the enlarged septum obstructs the aortic exit from the left ventricle. When pacing to reduce aortic obstruction, the contractions of the left and right ventricles are stepped, so that when the left ventricle contracts, the right ventricle dilates and the aorta is less compressed.
Lameh Fananapazir, Neal D. Epstein, Rodolfo V. Curiel, Julio A. Panza, Dorothy Tripodi and Dorothea McAreavey, in “Long-Term Results Of Dual-Chamber (DDD) Pacing In Obstructive Hypertrophic Cardiomyopathy”, Circulation, Vol. 90, No. 60, pp. 2731–2742, December 1994, the disclosure of which is incorporated herein by reference, describes the effects of pacing a HCM-diseased heart using DDD pacing at the apex of the right ventricle. One effect is that the muscle mass near the pacing location is reduced, i.e., the ventricular septum is atrophied. The atrophy is hypothesized to be caused by the changes in workload at the paced location which are due to the late activation time of ventricular segments far from the pacing location.
Margarete Hochleitner, Helmut Hortnagl, Heide Hortnagl, Leo Fridrich and Franz Gschnitzer, in “Long-Term Efficiency Of Physiologic Dual-Chamber Pacing In The Treatment Of End-Stage Idiopathic Dilated Cardiomyopathy”, American Journal of Cardiology, volume 70, pp. 1320–1325, 1992, the disclosure of which is incorporated herein by reference, describes the effect of DDD pacing on hearts which are dilated as a result of idiopathic dilated cardiomyopathy. DDD pacing resulted in an improvement of cardiac function and in a reduction in hypertrophy in several patients. In addition, it is suggested that positioning the ventricular electrode of the DDD pacemaker in near the apex of the right ventricle reduced the stress at the apex of the left ventricle, by its early activation. No method is suggested for choosing the implantation location of the electrodes.
Xavier Jeanrenaud, Jean-Jacques Goy and Lukas Kappenberger, in “Effects Of Dual Chamber Pacing In Hypertrophic Obstructive Cardiomyopathy”, The Lancet, Vol. 339, pp. 1318–1322, May 30, 1992, the disclosure of which is incorporated herein by reference, teaches that to ensure success of DDD pacing in HCM diseased hearts, an optimum AV interval (between atrial activation and ventricular activation) is required. In addition, it is suggested that this optimal AV interval is modified by performing exercise.
Several methods may be used to treat heart failure. One method is to connect assist pumps to the patient's circulatory system, which assist the heart by circulating the blood. To date, no satisfactory long-term assist pump has been developed. In some cases, a diseased heart is removed and replaced by another human heart. However, this is an expensive, complicated and dangerous operation and too few donor hearts are available. Artificial hearts suffer from the same limitations as assist pumps and, like them, are not yet practical.
Certain types of heart failure, such as those caused by conduction blocks in the AV node or by AF can be helped by the implantation of a pacemaker, as described above.
Some cases of heart failure can be helped by medicines which either strengthen the heart, correct arrhythmias or reduce the total volume of blood in the body (which reduces blood pressure): However, many cases of heart failure can only be treated by reducing the activity of the patient. Ultimately, once the cardiac reserve is used up, most cases of heart failure cannot be treated and result in death.
U.S. Pat. No. 5,391,199, the disclosure of which is incorporated herein by reference, discloses apparatus and method for mapping the electrical activity of the heart. “Biomedical Engineering Handbook”, ed. Joseph D. Bronzino, chapter 156.3, pp. 2371–2373, IEEE press/CRC press, 1995, describes modeling strategies in cardiac physiology. On page 2373 a model is described, including experimental support, according to which model the shape of a ventricle is determined by the (local) amount of oxygen consumption. In addition, this model differentiates between pressure overload on the heart, which causes thickening of muscle fibers, denoted concentric hypertrophy, and volume overload which causes an increase in the ventricular volume (by stretching), denoted eccentric hypertrophy. Eccentric hypertrophy may also be caused by reducing the amount of oxygen available to the cardiac muscle.
R. S. Reneman, F. W. Prinzen, E. C. Cheriex, T. Arts and T. Delhass, in “Asymmetrical Changes in Left Ventricular Diastolic Wall Thickness Induced by Chronic Asynchronous Electrical Activation in Man and Dogs”, FASEB J., 1993;7; A752 (abstract), abstract number 4341, the disclosure of which in incorporated herein by reference, describe results of studies in paced hearts and which show that earlier activated ventricular wall portions were thinner than later activated wall portions, showing an asymmetrical hypertrophy as a result of the pacing.
C. Daubert, PH. Mabo, Veronique Berder, D. Gras and C. LeClercq, in “Atrial Tachyarrhythmias Associated with High Degree Interatrial Conduction Block: Prevention by Permanent Atrial Resynchronisation”, European Journal of C.P.F, Vol. 4, No. 1, pp. 35–44, 1994, the disclosure of which is incorporated herein by reference, describes a method of treating atrial fibrillation by implanting pacemaker electrodes in various locations in the heart, including two electrodes in the right atrium.
Frits W. Prinzen, Cornelis H. Augustijn, Theo Arts, Maurits A. Allessie and Robert Reneman, in “Redistribution of Myocardial Fiber Strain and Blood Flow by Asynchronous Activation”, American Journal of Physiology No. 259 (Heart Circulation Physiology No. 28), H300–H308, 1990, the disclosure of which is incorporated herein by reference, describes studies which show that the location of pacing electrodes in a paced heart significantly affect the distribution of strain, and perfusion (blood flow) in the heart.